These apparently random remembrances, triggered by a sensory cue representing only a portion of the original memory, appear to be dependent on a particular region of the brain--the CA3 region of the hippocampus, say researchers from Baylor College of Medicine in Houston, collaborating with others from the Massachusetts Institute of Technology in Cambridge, and Hokkaido University School of Medicine in Sapporo Japan. Their findings appear in an article in the Journal Science, available currently (May 30, 2002) at the Science Express web site (www.sciencexpress.org).
"It appears that the CA3 region of the hippocampus is essential for the phenomenon called 'pattern completion'," said Dr. Dan Johnston, professor of neuroscience at Baylor College of Medicine. "That is the ability to recall memories from partial representations of the original."
The scientists genetically engineered a strain of mice in which a specific protein called the NMDA receptor was absent only in the CA3 region of
the hippocampus. This receptor senses the primary neurotransmitter of the brain, glutamate, in ways thought to be necessary for some forms of learning and memory. When the mutant animals were trained in a maze with several visible cues, they performed the task as well as normal animals. However, when some, but not all of the cues were removed, the mutant animals could not remember where they were supposed to go. The normal animals did fine.
"The hippocampus appears to be the place where some memories are initially processed and stored," said Johnston. He believes that those memories can remain there for up to a year before being transferred to other brain regions, specifically the cortex. It is known that a person whose hippocampus is injured cannot form new memories, but can access older ones that presumably reside in the cortex.
Johnston and researchers in his laboratory have measured the electrical activity mediated by the NMDA receptor in the hippocampus of mutant and normal mice to help their colleagues from other institutions demonstrate that a lack of NMDA receptor function within the CA3 region was responsible for the deficits displayed by the mutant animals.
"The next step is to determine how the neurons of the hippocampal CA3 region process the information to re-create a memory from only a portion of the original," said Johnston.
Others who participated in the research include Drs. Raymond A. Chitwood and Mark F. Yeckel of Baylor; Drs. Kazu Nakazawa, Michael C. Quirk, Linus D. Sun, Akira Kato, Candice Carr, Matthew A. Wilson and Susumu Tonegawa of MIT, and Dr. Masahiko Watanabe of Hokkaido University.